Miscellaneous

Code cleanup and modernization:

Use flake8? See also the configuration in ../../tox.ini.

Check and solve issue from PEP 290 - Code Migration and Modernization. (Covers issues up to Python 2.4, is there an equivalent for more recent modernizations?)

Ensure backwards compatibility!

• Improve handling on Windows:

  • Get graphical installer.
  • Make rst2html.py an .exe file using py2exe.

• The user interface is very difficult to use for most Windows users; you can't really expect them to use the command line. We need some kind of GUI that can launch rst2html.py, and save the HTML output to a file, and launch a browser. What's important is that we get settings to work with the GUI. So we need some way to dynamically generate a list of settings for the GUI. The current settings_spec for OptionParser doesn't seem to be usable for this for the following reasons:

  • It's biased toward the command line -- there are two options for one boolean setting.
  • You cannot have both a one-line description and a longer description for tooltips/help-texts.
  • It doesn't provide hints for the input type. You cannot easily infer the type of a setting from its validator, because any component can add new validators. In fact, it may be necessary to have both a hint about the input type (e.g. string) and a validator (valid ID), or it may be necessary to have a different set of choices for the CLI (1, INFO, 2, ...) and for the GUI (INFO, WARNING, ...).
  • It's coupled to the OptionParser. We want to be able to change the underlying system without breaking everything.
  • It's a bunch of primitive structures. We want an extensible (thus object-oriented) interface.

  So we probably need to create a class for storing all the settings, and auto-generate the OptionParser data from that.

  I talked to Stephan Deibel about getting Docutils integrated into Wing IDE. He said it's possible, and he'd be willing to help. There's a scripting interface to Wing, which we'd use. We can dynamically generate a list of preferences and not worry too much about the rendering (from what I understood); Wing's whole GUI is dynamic anyway. The interface could be made usable for other GUIs. For example, we could try to get option support for DocFactory. // FW

• Allow different report levels for STDERR and system_messages inside the document?

• Change the docutils-update script (in sandbox/infrastructure), to support arbitrary branch snapshots.

• Move some general-interest sandboxes out of individuals' directories, into subprojects?

• Add option for file (and URL) access restriction to make Docutils usable in Wikis and similar applications.

  2005-03-21: added file_insertion_enabled & raw_enabled settings. These partially solve the problem, allowing or disabling all file accesses, but not limited access.

• Configuration file handling needs discussion:

  • There should be some error checking on the contents of config files. How much checking should be done? How loudly should Docutils complain if it encounters an error/problem?
  • Docutils doesn't complain when it doesn't find a configuration file supplied with the --config option. Should it? (If yes, error or warning?)

• Internationalization:

  • I18n needs refactoring, the language dictionaries are difficult to maintain. Maybe have a look at gettext or similar tools.

    (This would make a nice Google Summer of Code project)

  • Language modules: in accented languages it may be useful to have both accented and unaccented entries in the bibliographic_fields mapping for versatility.

  • Add a "--strict-language" option & setting: no English fallback for language-dependent features.

    Make this the default for output (as opposed to input)? Throw an error with a helpful message, e.g.

    Default "contents" title for language %s missing, please specify an explicit title.

    or

    "attention" title for language %s missing, please use a generic admonition with explicit title.

  • Add internationalization to footer boilerplate text (resulting from "--generator", "--source-link", and "--date" etc.), allowing translations.

• Add validation? See http://pytrex.sourceforge.net, RELAX NG, pyRXP.

• In docutils.readers.get_reader_class (& parsers & writers too), should we be importing "standalone" or "docutils.readers.standalone"? (This would avoid importing top-level modules if the module name is not in docutils/readers. Potential nastiness.)

• Perhaps store a name-to-id mapping file? This could be stored permanently, read by subsequent processing runs, and updated with new entries. ("Persistent ID mapping"?)

• Perhaps the Component.supports method should deal with individual features ("meta" etc.) instead of formats ("html" etc.)? Currently, it is not used at all.

  Do we need it at all? Or rather let the writers just ignore some nodes (like we already do for "class" values)?

  The current implementation of the framework also leads to bug bug #241 "doctree-based publishing != publish_string". The "components.Filter" transform is run by publish_doctree(). When filtering based on the output format, it should be run by publish_from_doctree() instead because only then the writer is known.

  So we need to either remove or fix the framework.

• Think about large documents made up of multiple subdocument files. Issues: continuity (persistent sequences above), cross-references (name-to-id mapping file above and targets in other documents below), splitting (document splitting below).

  When writing a book, the author probably wants to split it up into files, perhaps one per chapter (but perhaps even more detailed). However, we'd like to be able to have references from one chapter to another, and have continuous numbering (pages and chapters, as applicable). Of course, none of this is implemented yet. There has been some thought put into some aspects; see the "include" directive and the Reference Merging transform below.

  When I was working with SGML in Japan, we had a system where there was a top-level coordinating file, book.sgml, which contained the top-level structure of a book: the <book> element, containing the book <title> and empty component elements (<preface>, <chapter>, <appendix>, etc.), each with filename attributes pointing to the actual source for the component. Something like this:

  <book id="bk01">
  <title>Title of the Book</title>
  <preface inrefid="pr01"></preface>
  <chapter inrefid="ch01"></chapter>
  <chapter inrefid="ch02"></chapter>
  <chapter inrefid="ch03"></chapter>
  <appendix inrefid="ap01"></appendix>
  </book>
  

  (The "inrefid" attribute stood for "insertion reference ID".)

  The processing system would process each component separately, but it would recognize and use the book file to coordinate chapter and page numbering, and keep a persistent ID to (title, page number) mapping database for cross-references. Docutils could use a similar system for large-scale, multipart documents.

  Aahz's idea:

  First the ToC:

  .. ToC-list::
      Introduction.txt
      Objects.txt
      Data.txt
      Control.txt
  

  Then a sample use:

  .. include:: ToC.txt
  
  As I said earlier in chapter :chapter:`Objects.txt`, the
  reference count gets increased every time a binding is made.
  

  Which produces:

  As I said earlier in chapter 2, the
  reference count gets increased every time a binding is made.
  

  The ToC in this form doesn't even need to be references to actual reST documents; I'm simply doing it that way for a minimum of future-proofing, in case I do want to add the ability to pick up references within external chapters.

  Perhaps, instead of ToC (which would overload the "contents" directive concept already in use), we could use "manifest". A "manifest" directive might associate local reference names with files:

  .. manifest::
     intro: Introduction.txt
     objects: Objects.txt
     data: Data.txt
     control: Control.txt
  

  Then the sample becomes:

  .. include:: manifest.txt
  
  As I said earlier in chapter :chapter:`objects`, the
  reference count gets increased every time a binding is made.
  

• Add support for multiple output files and generic data handling:

  It should be possible for a component to emit or reference data to be either included or referenced in the output document. Examples of such data are stylesheets or images.

  For this, we need a "data" object which stores the data either inline or by referring to a file. The Docutils framework is responsible for either:

  • storing the data in the appropriate location (e.g. in the directory of the output file, or in a user-specified directory) and providing the paths of the stored files to the writer, or
  • providing the data itself to the writer so that it can be embedded in the output document.

  This approach decouples data handling from the data source (which can either be embedded or referenced) and the destination (which can either be embedded or referenced as well).

  See <http://article.gmane.org/gmane.text.docutils.devel/3631>.

• Add testing for Docutils' front end tools?

• Publisher: "Ordinary setup" shouldn't require specific ordering; at the very least, there ought to be error checking higher up in the call chain. [Aahz]

  Publisher.get_settings requires that all components be set up before it's called. Perhaps the I/O objects shouldn't be set, but I/O classes. Then options are set up (.set_options), and Publisher.set_io (or equivalent code) is called with source & destination paths, creating the I/O objects.

  Perhaps I/O objects shouldn't be instantiated until required. For split output, the Writer may be called multiple times, once for each doctree, and each doctree should have a separate Output object (with a different path). Is the "Builder" pattern applicable here?

• Perhaps I/O objects should become full-fledged components (i.e. subclasses of docutils.Component, as are Readers, Parsers, and Writers now), and thus have associated option/setting specs and transforms.

• Multiple file I/O suggestion from Michael Hudson: use a file-like object or something you can iterate over to get file-like objects.

• Add an "--input-language" option & setting? Specify a different language module for input (bibliographic fields, directives) than for output. The "--language" option would set both input & output languages.

• Auto-generate reference tables for language-dependent features? Could be generated from the source modules. A special command-line option could be added to Docutils front ends to do this. (Idea from Engelbert Gruber.)

• Enable feedback of some kind from internal decisions, such as reporting the successful input encoding. Modify runtime settings? System message? Simple stderr output?

• Rationalize Writer settings (HTML/LaTeX/PEP) -- share settings.

• Add an "--include file" command-line option (config setting too?), equivalent to ".. include:: file" as the first line of the doc text? Especially useful for character entity sets, text transform specs, boilerplate, etc.

• Support "include" as embedded inline-compatible directive in substitution definitions, e.g.

  .. |version| include:: version.txt
  
  This document describes version |version| of ...
  

  (cf. Grzegorz Adam Hankiewicz's post from 2014-10-01 in docutils-devel)

• Add an :optional: <replacement text> option to the "include" directive? This would not throw an error for a missing file, instead a warning is given and <replacement text> is used instead. It would be the responsibility of the author to ensure the missing file does not lead to problems later in the document.

  Use cases:

  • Standard rST syntax to replace Sphinx's "literalinclude":

    .. include:: blah.cpp
       :literal:
       :optional: file ``blah.cpp`` not found
    

  • Variable content taken from a file, e.g.

    version.txt:

    .. |version| replace:: 3.1
    

    optionally used as:

    .. include:: version.txt
       :optional: .. |version| replace:: unknown
    
    This document describes version |version| of ...
    

    (cf. Grzegorz Adam Hankiewicz's post from 2014-10-01 in docutils-devel)

• Parameterize the Reporter object or class? See the 2004-02-18 "rest checking and source path" thread.

• Add a "disable_transforms" setting? Would allow for easy syntax checking. Where ("null" writer, generic, parser(s))? Cf. the 2004-02-18 "rest checking and source path" thread.

• Add a generic meta-stylesheet mechanism? An external file could associate style names ("class" attributes) with specific elements. Could be generalized to arbitrary output attributes; useful for HTML & XMLs. Aahz implemented something like this in sandbox/aahz/Effective/EffMap.py.

• William Dode suggested that table cells be assigned "class" attributes by columns, so that stylesheets can affect text alignment. Unfortunately, there doesn't seem to be a way (in HTML at least) to leverage the "colspec" elements (HTML "col" tags) by adding classes to them. The resulting HTML is very verbose:

  <td class="col1">111</td>
  <td class="col2">222</td>
  ...
  

  At the very least, it should be an option. People who don't use it shouldn't be penalized by increases in their HTML file sizes.

  Table rows could also be assigned classes (like odd/even). That would be easier to implement.

  How should it be implemented?

  • There could be writer options (column classes & row classes) with standard values.
  • The table directive could grow some options. Something like ":cell-classes: col1 col2 col3" (either must match the number of columns, or repeat to fill?) and ":row-classes: odd even" (repeat to fill; body rows only, or header rows too?).

  Probably per-table directive options are best. The "class" values could be used by any writer, and applying such classes to all tables in a document with writer options is too broad.

  See also the table_styling Sphinx extension which defines

  widths:	also in Docutils core (but different implementation)
  column-alignment:
  	Sets per-column text alignment
  column-wrapping:
  	Sets per-column text wrapping
  column-dividers:
  	Add dividers between columns
  column-classes:	Add per-column css classes.
  header-columns:	Specify number of “stub” columns

• Add file-specific settings support to config files, like:

  [file index.txt]
  compact-lists: no
  

  Is this even possible? Should the criterion be the name of the input file or the output file? Alternative (more explicit) syntax:

  [source_file index.txt]
  ...
  
  [dest_file index.html]
  ...
  

  Or rather allow settings configuration from the rst source file (see misc.settings directive)?

• The "validator" support added to OptionParser is very similar to "traits" in SciPy. Perhaps something could be done with them? (Had I known about traits when I was implementing docutils.frontend, I may have used them instead of rolling my own.)

• tools/buildhtml.py: Extend the --prune option ("prune" config setting) to accept file names (generic path) in addition to directories (e.g. --prune=docs/user/rst/cheatsheet.txt, which should not be converted to HTML).

• Add support for plugins.

• Config directories: Currently, ~/.docutils, ./docutils.conf/, & /etc/docutils.conf are read as configuration files. Proposal: allow ~/.docutils to be a a configuration directory, along with /etc/docutils/ and ./docutils.conf/. Within these directories, check for config.txt files. We can also have subdirectories here, for plugins, S5 themes, components (readers/writers/parsers) etc.

  Docutils will continue to support configuration files for backwards compatibility.

• Add support for document decorations other than headers & footers? For example, top/bottom/side navigation bars for web pages. Generic decorations?

  Seems like a bad idea as long as it isn't independent from the output format (for example, navigation bars are only useful for web pages).

• docutils_update: Check for a Makefile in a directory, and run make if found? This would allow for variant processing on specific source files, such as running rst2s5.py instead of rst2html.py.

• Add a "disable table of contents" setting? The S5 writer could set it as a default. Rationale:

  The contents (table of contents) directive must not be used [in S5/HTML documents]. It changes the CSS class of headings and they won't show up correctly in the screen presentation.

  —Easy Slide Shows With reStructuredText & S5

  Analogue to the sectnum_xform setting, it could be used by the latex writer to switch to a LaTeX generated ToC (currently, the latex writer calls it "use_latex_toc").

object numbering and object references

For equations, tables & figures.

These would be the equivalent of DocBook's "formal" elements.

In LaTeX, automatic counters are implemented for sections, equations and floats (figures, tables) (configurable via stylesheets or in the latex-preamble). Objects can be given reference names with the \label{<refname} command, \ref{<refname>} inserts the corresponding number.

No such mechanism exists in HTML.

• We need persistent sequences, similar to chapter and footnote numbers. See OpenOffice.org XML "fields".

  • Should the sequences be automatic or manual (user-specifyable)?

• It is already possible to give reference names to objects via internal hyperlink targets or the "name" directive option:

  .. _figure name:
  
  .. figure:: image.png
  

  or

  .. figure:: image.png
     :name: figure name
  

  Improve the mapping of "phrase references" to IDs/labels with Literal transcription (i.e. ü -> ue, ß -> ss, å -> aa) instead of just stripping the accents and other non-ASCII chars. See also the feature request allow more characters when transforming "names" to "ids".

  A "table" directive has been implemented, supporting table titles.

  Perhaps the name could derive from the title/caption?

• We need syntax for object references. Cf. OpenOffice.org XML "reference fields":

  • Parameterized substitutions are too complicated (cf. or not to do: object references)

  • An interpreted text approach is simpler and better:

    See Figure :ref:`figure name` and Equation :ref:`eq:identity`.
    

  • "equation", "figure", and "page" roles could generate appropriate boilerplate text:

    See :figure:`figure name` on :page:`figure name`.
    

    See Interpreted Text below.

    Reference boilerplate could be specified in the document (defaulting to nothing):

    .. fignum::
       :prefix-ref: "Figure "
       :prefix-caption: "Fig. "
       :suffix-caption: :
    

    The position of the role (prefix or suffix) could also be utilized

See also the Modified rst2html by Nicolas Rougier for a sample implementation.

User Docs

• Add a FAQ entry about using Docutils (with reStructuredText) on a server and that it's terribly slow. See the first paragraphs in <http://article.gmane.org/gmane.text.docutils.user/1584>.
• Add document about what Docutils has previously been used for (web/use-cases.txt?).
• Improve index in docs/user/config.txt.

Developer Docs

• Improve the internal module documentation (docstrings in the code). Specific deficiencies listed below.
  • docutils.parsers.rst.states.State.build_table: data structure required (including StringList).
  • docutils.parsers.rst.states: more complete documentation of parser internals.
• docs/ref/doctree.txt: DTD element structural relationships, semantics, and attributes. In progress; element descriptions to be completed.
• Document the pending elements, how they're generated and what they do.
• Document the transforms (perhaps in docstrings?): how they're used, what they do, dependencies & order considerations.
• Document the HTML classes used by html4css1.py.
• Write an overview of the Docutils architecture, as an introduction for developers. What connects to what, why, and how. Either update PEP 258 (see PEPs below) or as a separate doc.
• Give information about unit tests. Maybe as a howto?
• Document the docutils.nodes APIs.
• Complete the docs/api/publisher.txt docs.

How-Tos

• Creating Docutils Writers
• Creating Docutils Readers
• Creating Docutils Transforms
• Creating Docutils Parsers
• Using Docutils as a Library

PEPs

• Complete PEP 258 Docutils Design Specification.

  • Fill in the blanks in API details.

  • Specify the nodes.py internal data structure implementation?

    [Tibs:] Eventually we need to have direct documentation in there on how it all hangs together - the DTD is not enough (indeed, is it still meant to be correct? [Yes, it is. --DG]).

• Rework PEP 257, separating style from spec from tools, wrt Docutils? See Doc-SIG from 2001-06-19/20.

Bugs

• A container directive with :class: option gets the spurious class value "class".

Misc

• Another list problem:

  * foo
        * bar
        * baz
  

  This ends up as a definition list. This is more of a usability issue.

• This case is probably meant to be a nested list, but it ends up as a list inside a block-quote without an error message:

  - foo
  
   - bar
  

  It should probably just be an error.

  The problem with this is that you don't notice easily in HTML that it's not a nested list but a block-quote -- there's not much of a visual difference.

• Treat enumerated lists that are not arabic and consist of only one item in a single line as ordinary paragraphs. See <http://article.gmane.org/gmane.text.docutils.user/2635>.

• The citation syntax could use some improvements. See <http://thread.gmane.org/gmane.text.docutils.user/2499> (and the sub-thread at <http://thread.gmane.org/gmane.text.docutils.user/2499/focus=3028>, and the follow-ups at <http://thread.gmane.org/gmane.text.docutils.user/3087>, <http://thread.gmane.org/gmane.text.docutils.user/3110>, <http://thread.gmane.org/gmane.text.docutils.user/3114>), <http://thread.gmane.org/gmane.text.docutils.user/2443>, <http://thread.gmane.org/gmane.text.docutils.user/2715>, <http://thread.gmane.org/gmane.text.docutils.user/3027>, <http://thread.gmane.org/gmane.text.docutils.user/3120>, <http://thread.gmane.org/gmane.text.docutils.user/3253>.

• The current list-recognition logic has too many false positives, as in

  * Aorta
  * V. cava superior
  * V. cava inferior
  

  Here V. is recognized as an enumerator, which leads to confusion. We need to find a solution that resolves such problems without complicating the spec to much.

  See <http://thread.gmane.org/gmane.text.docutils.user/2524>.

• Add indirect links via citation references & footnote references. Example:

  `Goodger (2005)`_ is helpful.
  
  .. _Goodger (2005): [goodger2005]_
  .. [goodger2005] citation text
  

  See <http://thread.gmane.org/gmane.text.docutils.user/2499>.

• Complain about bad URI characters (http://article.gmane.org/gmane.text.docutils.user/2046) and disallow internal whitespace (http://article.gmane.org/gmane.text.docutils.user/2214).

• Create info-level system messages for unnecessarily backslash-escaped characters (as in "\something", rendered as "something") to allow checking for errors which silently slipped through.

• Add (functional) tests for untested roles.

• Add test for ":figwidth: image" option of "figure" directive. (Test code needs to check if PIL is available on the system.)

• Add support for CJK double-width whitespace (indentation) & punctuation characters (markup; e.g. double-width "*", "-", "+")?

• Add motivation sections for constructs in spec.

• Support generic hyperlink references to targets in other documents? Not in an HTML-centric way, though (it's trivial to say https://www.example.org/doc#name, and useless in non-HTML contexts). XLink/XPointer? .. baseref::? See Doc-SIG 2001-08-10.

• Implement the header row separator modification to table.el. (Wrote to Takaaki Ota & the table.el mailing list on 2001-08-12, suggesting support for "=====" header rows. On 2001-08-17 he replied, saying he'd put it on his to-do list, but "don't hold your breath".)

• Fix the parser's indentation handling to conform with the stricter definition in the spec. (Explicit markup blocks should be strict or forgiving?)

• Make the parser modular. Allow syntax constructs to be added or disabled at run-time. Subclassing is probably not enough because it makes it difficult to apply multiple extensions.

• Generalize the "doctest block" construct (which is overly Python-centric) to other interactive sessions? "Doctest block" could be renamed to "I/O block" or "interactive block", and each of these could also be recognized as such by the parser:

  • Shell sessions:

    $ cat example1.txt
    A block beginning with a "$ " prompt is interpreted as a shell
    session interactive block.  As with Doctest blocks, the
    interactive block ends with the first blank line, and wouldn't
    have to be indented.
    

  • Root shell sessions:

    # cat example2.txt
    A block beginning with a "# " prompt is interpreted as a root
    shell session (the user is or has to be logged in as root)
    interactive block.  Again, the block ends with a blank line.
    

  Other standard (and unambiguous) interactive session prompts could easily be added (such as "> " for WinDOS).

  Tony Ibbs spoke out against this idea (2002-06-14 Doc-SIG thread "docutils feedback").

• Add support for pragma (syntax-altering) directives.

  Some pragma directives could be local-scope unless explicitly specified as global/pragma using ":global:" options.

• Support whitespace in angle-bracketed standalone URLs according to Appendix E ("Recommendations for Delimiting URI in Context") of RFC 2396.

• Use the vertical spacing of the source text to determine the corresponding vertical spacing of the output?

• [From Mark Nodine] For cells in simple tables that comprise a single line, the justification can be inferred according to the following rules:

  1. If the text begins at the leftmost column of the cell, then left justification, ELSE
  2. If the text begins at the rightmost column of the cell, then right justification, ELSE
  3. Center justification.

  The onus is on the author to make the text unambiguous by adding blank columns as necessary. There should be a parser setting to turn off justification-recognition (normally on would be fine).

  Decimal justification?

  All this shouldn't be done automatically. Only when it's requested by the user, e.g. with something like this:

  .. table::
     :auto-indent:
  
     (Table goes here.)
  

  Otherwise it will break existing documents.

• Generate a warning or info message for paragraphs which should have been lists, like this one:

  1. line one
  3. line two
  

• Generalize the "target-notes" directive into a command-line option somehow? See docutils-develop 2003-02-13.

• Allow a "::"-only paragraph (first line, actually) to introduce a literal block without a blank line? (Idea from Paul Moore.)

  ::
      This is a literal block
  

  Is indentation enough to make the separation between a paragraph which contains just a :: and the literal text unambiguous? (There's one problem with this concession: If one wants a definition list item which defines the term "::", we'd have to escape it.) It would only be reasonable to apply it to "::"-only paragraphs though. I think the blank line is visually necessary if there's text before the "::":

  The text in this paragraph needs separation
  from the literal block following::
      This doesn't look right.
  

• Add new syntax for nested inline markup? Or extend the parser to parse nested inline markup somehow? See the collected notes.

• Drop the backticks from embedded URIs with omitted reference text? Should the angle brackets be kept in the output or not?

  <file_name>_
  

  Probably not worth the trouble.

• How about a syntax for alternative hyperlink behavior, such as "open in a new window" (as in HTML's <a target="_blank">)?

  The MoinMoin wiki uses a caret ("^") at the beginning of the URL ("^" is not a legal URI character). That could work for both inline and explicit targets:

  The `reference docs <^url>`__ may be handy.
  
  .. _name: ^url
  

  This may be too specific to HTML. It hasn't been requested very often either.

• Add an option to add URI schemes at runtime.

• Segmented lists:

  : segment : segment : segment
  : segment : segment : very long
    segment
  : segment : segment : segment
  

  The initial colon (":") can be thought of as a type of bullet

  We could even have segment titles:

  :: title  : title   : title
  : segment : segment : segment
  : segment : segment : segment
  

  This would correspond well to DocBook's SegmentedList. Output could be tabular or "name: value" pairs, as described in DocBook's docs.

• Enable grid tables inside XML comments, where "--" ends comments.

  Implementation possibilities:

  1. Make the table syntax characters into "table" directive options. This is the most flexible but most difficult, and we probably don't need that much flexibility.
  2. Substitute "~" for "-" with a specialized directive option (e.g. ":tildes:").
  3. Make the standard table syntax recognize "~" as well as "-", even without a directive option. Individual tables would have to be internally consistent.
  4. Allow Unicode box characters for table markup (feature request [6])

  Directive options are preferable to configuration settings, because tables are document-specific. A pragma directive would be another approach, to set the syntax once for a whole document.

  Unicode box character markup would kill two birds with one stone.

  In the meantime, the list-table directive is a good replacement for grid tables inside XML comments.

• Generalize docinfo contents (bibliographic fields): remove specific fields, and have only a single generic "field"?

• Line numbers and "source" in system messages:

  • Add "source" and "line" keyword arguments to all Reporter calls? This would require passing source/line arguments along all intermediate functions (where currently only line is used).

    Or rather specify "line" only if actually needed?

    Currently, document.reporter uses a state machine instance to determine the "source" and "line" info from statemachine.input_lines if not given explicitly. Except for special cases, the "line" argument is not needed because, document.statemachine keeps record of the current line number.

  • For system messages generated after the parsing is completed (i.e. by transforms or the writer) "line" info must be present in the doctree elements.

    Elements' .line assignments should be checked. (Assign to .source too? Add a set_info method? To what?)

    The "source" (and line number in the source) can either be added explicitly to the elements or determined from the “raw” line number by document.statemachine.get_source_and_line.

  • Some line numbers in elements are not being set properly (explicitly), just implicitly/automatically. See rev. 1.74 of docutils/parsers/rst/states.py for an example of how to set.

  • The line numbers of definition list items are wrong:

    $ rst2pseudoxml.py --expose-internal-attribute line
    1
      2
      3
    
    5
      6
      7
    
    <document source="<stdin>">
        <definition_list>
            <definition_list_item internal:line="3">
                <term>
                    1
                <definition>
                    <paragraph internal:line="2">
                        2
                        3
            <definition_list_item internal:line="6">
                <term>
                    5
                <definition>
                    <paragraph internal:line="6">
                        6
                        7
    

• Quite a few nodes are getting a "None" source attribute as well. In particular, see the bodies of definition lists.

Adaptable file extensions

.. _Another Document: another.*

  .. role:: link(rewrite)
     :transform: .txt|.html

and then to use it::

  for more information see :link:`README.txt`

it would be useful if it supported an additional option
``:format: html`` so that separate rules for each format can be
defined. (like the "raw" role)

rst2html --map-extension rst html --map-extension jpg png \
   input-filename.rst

Math Markup

Since Docutils 0.8, a "math" role and directive using LaTeX math syntax as input format is part of reStructuredText.

Open issues:

• Use a "Transform" for math format conversions as extensively discussed in the "math directive issues" thread in May 2008 (http://osdir.com/ml/text.docutils.devel/2008-05/threads.html)?
• Generic math-output setting (currently specific to HTML). (List of math-output preferences?)
• Try to be compatible with Math support in Sphinx?
  • The :label: option selects a label for the equation, by which it can be cross-referenced, and causes an equation number to be issued. In Docutils, the option :name: sets the label. Equation numbering is not implemented yet.
  • Option :nowrap: prevents wrapping of the given math in a math environment (you have to specify the math environment in the content).
• Equation numbering and references. (see the section on object numbering and object references for equations, formal tables, and images.)

Directives

Directives below are often referred to as "module.directive", the directive function. The "module." is not part of the directive name when used in a document.

• Allow for field lists in list tables. See <http://thread.gmane.org/gmane.text.docutils.devel/3392>.

• Unify table implementations and unify options of table directives (http://article.gmane.org/gmane.text.docutils.user/1857).

• Allow directives to be added at run-time?

• Use the language module for directive option names?

• Add "substitution_only" and "substitution_ok" function attributes, and automate context checking?

• Implement options or features on existing directives:

  • All directives that produce titled elements should grow implicit reference names based on the titles.

  • Allow the :trim: option for all directives when they occur in a substitution definition, not only the unicode directive.

  • Add the "class" option to the unicode directive. For example, you might want to get characters or strings with borders around them.

  • images.figure: "title" and "number", to indicate a formal figure?

  • parts.sectnum: "local"?, "refnum"

    A "local" option could enable numbering for sections from a certain point down, and sections in the rest of the document are not numbered. For example, a reference section of a manual might be numbered, but not the rest. OTOH, an all-or-nothing approach would probably be enough.

    The "sectnum" directive should be usable multiple times in a single document. For example, in a long document with "chapter" and "appendix" sections, there could be a second "sectnum" before the first appendix, changing the sequence used (from 1,2,3... to A,B,C...). This is where the "local" concept comes in. This part of the implementation can be left for later.

    A "refnum" option (better name?) would insert reference names (targets) consisting of the reference number. Then a URL could be of the form http://host/document.html#2.5 (or "2-5"?). Allow internal references by number? Allow name-based and number-based ids at the same time, or only one or the other (which would the table of contents use)? Usage issue: altering the section structure of a document could render hyperlinks invalid.

  • parts.contents: Add a "suppress" or "prune" option? It would suppress contents display for sections in a branch from that point down. Or a new directive, like "prune-contents"?

    Add an option to include topics in the TOC? Another for sidebars? The "topic" directive could have a "contents" option, or the "contents" directive" could have an "include-topics" option. See docutils-develop 2003-01-29.

  • parts.header & parts.footer: Support multiple, named headers & footers? For example, separate headers & footers for odd, even, and the first page of a document.

    This may be too specific to output formats which have a notion of "pages".

  • misc.class:

    • Add a :parent: option for setting the parent's class (http://article.gmane.org/gmane.text.docutils.devel/3165).

  • misc.include:

    • Option to label lines?

    • How about an environment variable, say RSTINCLUDEPATH or RSTPATH, for standard includes (as in .. include:: <name>)? This could be combined with a setting/option to allow user-defined include directories.

    • Add support for inclusion by URL?

      .. include::
         :url: https://www.example.org/inclusion.txt
      

    • Strip blank lines from begin and end of a literal included file or file section. This would correspond to the way a literal block is handled.

      As nodes.literal_block expects (and we have) the text as a string (rather than a list of lines), using a regexp seems the way.

  • misc.raw: add a "destination" option to the "raw" directive?

    .. raw:: html
       :destination: head
    
       <link ...>
    

    It needs thought & discussion though, to come up with a consistent set of destination labels and consistent behavior.

    And placing HTML code inside the <head> element of an HTML document is rather the job of a templating system.

  • body.sidebar: Allow internal section structure? Adornment styles would be independent of the main document.

    That is really complicated, however, and the document model greatly benefits from its simplicity.

• Implement directives. Each of the list items below begins with an identifier of the form, "module_name.directive_function_name". The directive name itself could be the same as the directive_function_name, or it could differ.

  • html.imagemap

    It has the disadvantage that it's only easily implementable for HTML, so it's specific to one output format.

    (For non-HTML writers, the imagemap would have to be replaced with the image only.)

  • parts.endnotes (or "footnotes"): See Footnote & Citation Gathering.

  • parts.citations: See Footnote & Citation Gathering.

  • misc.language: Specify (= change) the language of a document at parse time?

    • The misc.settings directive suggested below offers a more generic approach.
    • The language of document parts can be indicated by the "special class value" "language-" + BCP 47 language code. Class arguments to the title are attached to the document's base node - hence titled documents can be given a different language at parse time. However, "language by class attribute" does not change parsing (localized directives etc.), only supporting writers.

  • misc.settings: Set any(?) Docutils runtime setting from within a document? Needs much thought and discussion.

    Security concerns need to be taken into account (it shouldn't be possible to enable file_insertion_enabled from within a document), and settings that only would have taken effect before the directive (like tab-width) shouldn't be accessible either.

    See this sub-thread: <http://thread.gmane.org/gmane.text.docutils.user/3620/focus=3649>

  • misc.gather: Gather (move, or copy) all instances of a specific element. A generalization of the Footnote & Citation Gathering ideas.

  • Add a custom "directive" directive, equivalent to "role"? For example:

    .. directive:: incr
    
       .. class:: incremental
    
    .. incr::
    
    "``.. incr::``" above is equivalent to "``.. class:: incremental``".
    

    Another example:

    .. directive:: printed-links
    
       .. topic:: Links
          :class: print-block
    
          .. target-notes::
             :class: print-inline
    

    This acts like macros. The directive contents will have to be evaluated when referenced, not when defined.

    • Needs a better name? "Macro", "substitution"?
    • What to do with directive arguments & options when the macro/directive is referenced?

  • Make the meaning of block quotes overridable? Only a 1-shot though; doesn't solve the general problem.

  • conditional directives:

    Note

    See also the implementation in Sphinx.

    Docutils already has the ability to say "use this content for Writer X" via the "raw" directive. It also does have the ability to say "use this content for any Writer other than X" via the "strip-elements with class" config value. However, using "raw" input just to select a special writer is inconvenient in many cases. It wouldn't be difficult to get more straightforward support, though.

    My first idea would be to add a set of conditional directives. Let's call them "writer-is" and "writer-is-not" for discussion purposes (don't worry about implementation details). We might have:

    .. writer-is:: text-only
    
       ::
    
           +----------+
           |   SNMP   |
           +----------+
           |   UDP    |
           +----------+
           |    IP    |
           +----------+
           | Ethernet |
           +----------+
    
    .. writer-is:: pdf
    
       .. figure:: protocol_stack.eps
    
    .. writer-is-not:: text-only pdf
    
       .. figure:: protocol_stack.png
    

    This could be an interface to the Filter transform (docutils.transforms.components.Filter).

    The ideas in the adaptable file extensions section above may also be applicable here.

    SVG's "switch" statement may provide inspiration.

    Here's an example of a directive that could produce multiple outputs (both raw troff pass-through and a GIF, for example) and allow the Writer to select.

    .. eqn::
    
       .EQ
       delim %%
       .EN
       %sum from i=o to inf c sup i~=~lim from {m -> inf}
       sum from i=0 to m sup i%
       .EQ
       delim off
       .EN
    

  • body.example: Examples; suggested by Simon Hefti. Semantics as per Docbook's "example"; admonition-style, numbered, reference, with a caption/title.

  • body.index: Index targets.

    See Index Entries & Indexes.

  • body.literal: Literal block, possibly "formal" (see object numbering and object references above). Possible options:

    • "highlight" a range of lines

    • include only a specified range of lines

    • "number" or "line-numbers"? (since 0.9 available with "code" directive)

    • "styled" could indicate that the directive should check for style comments at the end of lines to indicate styling or markup.

      Specific derivatives (i.e., a "python-interactive" directive) could interpret style based on cues, like the ">>> " prompt and "input()"/"raw_input()" calls.

    See docutils-users 2003-03-03.

  • body.listing: Code listing with title (to be numbered eventually), equivalent of "figure" and "table" directives.

  • pysource.usage: Extract a usage message from the program, either by running it at the command line with a --help option or through an exposed API. [Suggestion for Optik.]

  • body.float: Generic float that can be used for figures, tables, code listings, flowcharts, ...

    There is a Sphinx extension by Ignacio Fernández Galván <jellby@gmail.com>

    I implemented something for generic floats in sphinx, and submitted a pull request that is still waiting:

    .. float::
       :type: figure
       :caption: My caption
    

    https://github.com/sphinx-doc/sphinx/pull/1858

Interpreted Text

Interpreted text is entirely a reStructuredText markup construct, a way to get around built-in limitations of the medium. Some roles are intended to introduce new doctree elements, such as "title-reference". Others are merely convenience features, like "RFC".

All supported interpreted text roles must already be known to the Parser when they are encountered in a document. Whether pre-defined in core/client code, or in the document, doesn't matter; the roles just need to have already been declared. Adding a new role may involve adding a new element to the DTD and may require extensive support, therefore such additions should be well thought-out. There should be a limited number of roles.

The only place where no limit is placed on variation is at the start, at the Reader/Parser interface. Transforms are inserted by the Reader into the Transformer's queue, where non-standard elements are converted. Once past the Transformer, no variation from the standard Docutils doctree is possible.

An example is the Python Source Reader, which will use interpreted text extensively. The default role will be "Python identifier", which will be further interpreted by namespace context into <class>, <method>, <module>, <attribute>, etc. elements (see pysource.dtd), which will be transformed into standard hyperlink references, which will be processed by the various Writers. No Writer will need to have any knowledge of the Python-Reader origin of these elements.

• Add explicit interpreted text roles for the rest of the implicit inline markup constructs: named-reference, anonymous-reference, footnote-reference, citation-reference, substitution-reference, target, uri-reference (& synonyms).

• Add directives for each role as well? This would allow indirect nested markup:

  This text contains |nested inline markup|.
  
  .. |nested inline markup| emphasis::
  
     nested ``inline`` markup
  

• Implement roles:

  • "raw-wrapped" (or "raw-wrap"): Base role to wrap raw text around role contents.

    For example, the following reStructuredText source ...

    .. role:: red(raw-formatting)
       :prefix:
           :html: <font color="red">
           :latex: {\color{red}
       :suffix:
           :html: </font>
           :latex: }
    
    colored :red:`text`
    

    ... will yield the following document fragment:

    <paragraph>
        colored
        <inline classes="red">
            <raw format="html">
                <font color="red">
            <raw format="latex">
                {\color{red}
            <inline classes="red">
                text
            <raw format="html">
                </font>
            <raw format="latex">
                }
    

    Possibly without the intermediate "inline" node.

  • "acronym" and "abbreviation": Associate the full text with a short form. Jason Diamond's description:

    I want to translate `reST`:acronym: into <acronym title='reStructuredText'>reST</acronym>. The value of the title attribute has to be defined out-of-band since you can't parameterize interpreted text. Right now I have them in a separate file but I'm experimenting with creating a directive that will use some form of reST syntax to let you define them.

    Should Docutils complain about undefined acronyms or abbreviations?

    What to do if there are multiple definitions? How to differentiate between CSS (Content Scrambling System) and CSS (Cascading Style Sheets) in a single document? David Priest responds,

    The short answer is: you don't. Anyone who did such a thing would be writing very poor documentation indeed. (Though I note that somewhere else in the docs, there's mention of allowing replacement text to be associated with the abbreviation. That takes care of the duplicate acronyms/abbreviations problem, though a writer would be foolish to ever need it.)

    How to define the full text? Possibilities:

    1. With a directive and a definition list?

       .. acronyms::
       
          reST
              reStructuredText
          DPS
              Docstring Processing System
       

       Would this list remain in the document as a glossary, or would it simply build an internal lookup table? A "glossary" directive could be used to make the intention clear. Acronyms/abbreviations and glossaries could work together.

       Then again, a glossary could be formed by gathering individual definitions from around the document.

    2. Some kind of inline parameter syntax?

       `reST <reStructuredText>`:acronym: is `WYSIWYG <what you
       see is what you get>`:acronym: plaintext markup.
       

    3. A combination of 1 & 2?

       The multiple definitions issue could be handled by establishing rules of priority. For example, directive-based lookup tables have highest priority, followed by the first inline definition. Multiple definitions in directive-based lookup tables would trigger warnings, similar to the rules of implicit hyperlink targets.

    4. Using substitutions?

       .. |reST| acronym:: reST
          :text: reStructuredText
       

    What do we do for other formats than HTML which do not support tool tips? Put the full text in parentheses?

  • "figure", "table", "listing", "chapter", "page", etc: See object numbering and object references above.

  • "glossary-term": This would establish a link to a glossary. It would require an associated "glossary-entry" directive, whose contents could be a definition list:

    .. glossary-entry::
    
       term1
           definition1
       term2
           definition2
    

    This would allow entries to be defined anywhere in the document, and collected (via a "glossary" directive perhaps) at one point.

Doctree pruning

[DG 2017-01-02: These are not definitive to-dos, just one developer's opinion. Added 2009-10-13 by Günter Milde, in r6178.] [Updated by GM 2017-02-04]

The number of doctree nodes can be reduced by "normalizing" some related nodes. This makes the document model and the writers somewhat simpler.

• The "doctest" element can be replaced by literal blocks with a class attribute (similar to the "code" directive output). The syntax shall be left in reST.

  [DG 2017-01-02:] +0.

  Discussion

  The syntax could be left in reST (for a set period of time?).

  [DG 2017-01-02:] The syntax must be left in reST, practically forever. Removing it would introduce a huge backwards incompatibility. Any syntax removal must be preceded by a thorough review and planning, including a deprecation warning process. My opinion: it's not worth it.

• "Normalize" special admonitions (note, hint, warning, ...) during parsing (similar to transforms.writer_aux.Admonitions). There is no need to keep them as distinct elements in the doctree specification.

  [DG 2017-01-02:] -1: <note>{body}</> is much more concise and expressive than <admonition><title>Note</>{body}</>, and the title translation can be put off until much later in the process.

  [GM 2017-02-04]:

  -0 for <admonition class=note><title>Note</>... instead of <note>:

  a document is rarely printed/used as doctree or XML.

  +1 reduce the complexity of the doctree

  (there is no 1:1 rST syntax element <-> doctree node mapping anyway).

  +2 every writer needs 9 visit_*/depart_* method pairs to handle the 9

  subtypes of an admonition, i.e. we could but also remove 36 redundant methods (HTML, LaTeX, Manpage, ODF).

  -1 the most unfortunately named of these directives will survive. [1]

  [1]	with "biblical touch" and hard to translate: admonition: Ermahnung; Verweis; Warnung; Rüge (exhortation; censure; warning; reprimand, rebuke)

  Keep the special admonition directives in reStructuredText syntax.

  [DG 2017-01-02:] We must definitely keep the syntax. Removing it would introduce a huge backwards incompatibility.

Bug fixes

• Too deeply nested lists fail: generate a warning and provide a workaround.

  2017-02-09 this is fixed for enumeration in 0.13.1

  for others, cf. sandbox/latex-variants/tests/rst-levels.txt

• File names of included graphics (see also grffile package).

• Paragraph following field-list or table in compound is indented.

  This is a problem with the current DUfieldlist definition and with the use of "longtable" for tables. See other LaTeX constructs and packages instead of re-implementations for alternatives.

Generate clean and configurable LaTeX source

Which packages do we want to use?

• base and "recommended" packages

  (packages that should be in a "reasonably sized and reasonably modern LaTeX installation like the texlive-latex-recommended Debian package, say):

• No "fancy" or "exotic" requirements.

• pointers to advanced packages and their use in the latex writer documentation.

Default layout

• Use italic instead of slanted for titlereference?

• Start a new paragraph after lists (as currently) or continue (no blank line in source, no parindent in output)?

  Overriding:

  • continue if the compound paragraph directive is used (as currently), or
  • force a new paragraph with an empty comment.

• Sidebar handling (environment with framed, marginnote, wrapfig, ...)?

• Use optionlist for docinfo?

• Keep literal-blocks together on a page, avoid pagebreaks.

  Failed experiments up to now: samepage, minipage, pagebreak 1 to 4 before the block.

  Should be possible with --literal-block-env==lstlistings and some configuration...

• More space between title and subtitle?

  -  \\ % subtitle%
  +  \\[0.5em] % subtitle%
  

Missing features

• support "figwidth" argument for figures.

  As the 'figwidth' argument is still ignored and the "natural width" of a figure in LaTeX is 100 % of the text width, setting the 'align' argument has currently no effect on the LaTeX output.

• Multiple author entries in docinfo (same thing as in html).

• Consider supporting the "compact" option and class argument (from rst2html) as some lists look better compact and others need the space.

• Better citation support (see Footnote & Citation Gathering).

• If use-latex-citations is used, a bibliography is inserted right at the end of the document.

  Put in place of the to-be-implemented "citations" directive (see Footnote & Citation Gathering).

\href{https://www.w3.org/XML/Schema^^dev}{\fragileURLi}